xml-dev - Re: Future of Formatting Objects (XSL/FO)

[Rick JELLIFFE <ricko@geotempo.com>]

Sebastian Rahtz wrote:
> 
> Peter Murray-Rust writes:
>  > So why can't FOP have the same dynamics? If a group of committed people
>  > want to drive it forward, what is lost?
> 
> I don't think anyone suggested that the FOP effort is anything other
> than very laudauble and useful. The point at issue is whether XSL FO
> is _theoretically_ capable of producing the high-quality formatting we
> are used to. We pretty sure that SVG is as least as capabable as
> PDF, because we can see the mapping between them. Half the attraction
> of SVG is that is very clearly related to highly-successful vector formats.
> I would claim that the jury is out on XSL FO.
 
Different criteria of quality are appropriate for different products. 

Knuth made an interesting comment about TeX that there will always be
the need to hand-tune some pages. For example, if there is a large
anchored graphic that breaks onto the next page, causing large amounts
of whitespace on the previous page. Or if a chapter ends with a couple
of lines on a recto (odd) page.

There are various approaches to overcome this: playing with line-length,
font-size, leading or column-lengths (e.g. Word Perfect's copyfitting);
providing warnings of pages that fail to meet minimum badness (e.g.
TeX); having smarter macro languages that can attempt some custom
fiddles (e.g. Interleaf).  

The approach I think would be nice is to be able to mark up alternative
portions, weighting them for "fluff" value. This would be especially
useful for newspapers.
So you could have

<p><span fluff="medium">American President</span> Bill 
Clinton<span fliff="high">, in town to support a capital
punishment amendment,</span> was rescued by Linda Tripp
from an attack by a disabled transvestite fisherman from
Florida<span fluff="medium"> outside the former OJ Simpson
estate</span>.</p>

I am excited to see the results of XSL-FO: it is looking great. But a
good page design will often need to be tuned to fit the formatter: a
formatter that does not provide feathering and which tries to keep
paragaphs together by pushing them forward a page (i.e., by its default
widow/orphan settings) will probably benefit from conservative amounts
of bottom margin, to reduce the whitespace at the bottom of pages.

May I also defend FrameMaker?  I have found it so convenient, especially
for producing PDF with the built-in Acrobat distiller. The thing I was
surprised about was that I even found myself making slides for seminars
using the structured editor: it was simply more convenient than using
WYSIWYG or vi or an HTML editor (my normal choices).  For CJK
publishing, it supports ruby characters. For complex XML/SGML, one can
do transformations outside and import, just as one would end up doing
with TeX, Word, WP, etc.  It has lots of nice tunable parameters: very
nice. For production work it is great. (If only it had as good a grammar
checker as Word Perfect's!)

Rick Jelliffe

***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@xml.org&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/
***************************************************************************